Shi Ding
Ph.D. Candidate · Georgia Institute of Technology
Expressive Machinery Lab · Advised by Dr. Brian Magerko
About
I’m Shi Ding (Stella), Ph.D. researcher at Georgia Tech focused on human-centered and responsible AI. I have extensive research experience designing and evaluating LLM-based agents for large-scale educational platforms. Since 2023, I have been a key contributor to the NSF-funded EarSketch research project in the Expressive Machinery Lab, advised by Dr. Brian Magerko, helping scale the platform to over 1.5 million learners globally.
Highlights & Achievements
- Presentation: Ding et al. (2025).Rethinking AI Evaluation with TEACH-AI: A Benchmark and Toolkit for Educational AI Systems, Accepted to the NeurIPS ResponsibleFM workshop (2025)
- Media Feature: “An AI Ally for Teachers”and “Rethinking AI in Learning: Inclusive Design, Responsible Impact,” featured by TSQATL Media.(2025).
- Invited: Disability and Access in Tech and AI Summit at Cornell University & Cornell Tech (2025).
- Reviewer: NeurIPS (2025), SIGCSE (2026), and Cognitive Science Society (2025).
- Publication: Smith, Ding, et al. (2025). Using co-design to investigate affordances of an expressive CS learning environment for BVI learners. Accepted to ACM ASSETS 2025.
- Publication: Ding et al. (2025). Considering Large Language Model Integrations in Expressive CS Learning Environments for BVI Learners Through Co-design. Accepted to AIED 2025.
- Publication: Ding et al. (2024). Redesigning EarSketch for Inclusive CS Education: A Participatory Design Approach. Presented at ACM IDC 2024 and CHI Workshop 2024.
- Co-founder & Awardee: Deeproject AI, Quadrant-i Startup Grant, Georgia Tech (2025 summer).
- Poster Judge: InterfaceNeuro Symposium, Georgia Tech (2025).
- Passed Ph.D. Qualifying Exam: Georgia Tech (2025).
- Winner: CRIDC Poster Competition, Georgia Tech (2025): “Promoting Creativity of BVI Learners in EarSketch Through Co-Design.”
- Research Intern: Graspy AI (2024).
- Best Project Award: CS 6460 Educational Technology, Georgia Tech (2024).